6.451 Principles of Digital Communication II, Spring 2005
- Publication date
- coding techniques, the Shannon limit of additive white Gaussian noise channels, performance analysis, Small signal constellations, performance analysis, coding gain, Hard-decision and soft-decision decoding, Introduction to binary linear block codes, Reed-Muller codes, finite fields, Reed-Solomon and BCH codes, binary linear convolutional codes, Viterbi and BCJR algorithms, Trellis representations of binary linear block codes, trellis-based ML decoding, Codes on graphs, sum-product, max-product, decoding algorithms, Turbo codes, LDPC codes and RA codes, Coding for the bandwidth-limited regime, Lattice codes., Trellis-coded modulation, Multilevel coding, Shaping
Instructor: Prof. David Forney
This course is the second of a two-term sequence with 6.450. The focus is on coding techniques for approaching the Shannon limit of additive white Gaussian noise (AWGN) channels, their performance analysis, and design principles. After a review of 6.450 and the Shannon limit for AWGN channels, the course begins by discussing small signal constellations, performance analysis and coding gain, and hard-decision and soft-decision decoding. It continues with binary linear block codes, Reed-Muller codes, finite fields, Reed-Solomon and BCH codes, binary linear convolutional codes, and the Viterbi algorithm.
More advanced topics include trellis representations of binary linear block codes and trellis-based decoding; codes on graphs; the sum-product and min-sum algorithms; the BCJR algorithm; turbo codes, LDPC codes and RA codes; and performance of LDPC codes with iterative decoding. Finally, the course addresses coding for the bandwidth-limited regime, including lattice codes, trellis-coded modulation, multilevel coding and shaping. If time permits, it covers equalization of linear Gaussian channels.
* NOTE: The Real Media files are the originals. The MP4s are transcoded versions for the iPod.
- 2009-04-25 11:12:39
Uploaded by MIT-OCW on